Proteins are ubiquitous in all living organisms, having definite functions that are related to their molecular structures in ways that are often still mysterious. Many diseases are the result of defective protein structure and drugs designed to control disease may have as their target specific small-molecule binding sites on proteins. Understanding the details of protein structure and protein-small molecule complexes, and the dynamical properties of these systems, should lead to a better understanding of protein function and provide principles to guide the design of more effective pharmaceutical interventions for the control of disease. Nuclear magnetic resonance (nmr) spectroscopy is emerging as an important tool for the elucidation of protein structures in solution and for studying the binding of small molecules to proteins. These experiments are less successful as the size of the protein increases. Several research groups, including this laboratory, have shown that selective introduction of fluorine nuclei into a protein provides a way to make nmr techniques productive with larger proteins. However, there are experimental and interpretational aspects of this approach that are not adequately understood at present. These include the following. (1) Three-dimensional structural information is usually inferred from interpretation of nmr relaxation data but the reliability of these conclusions is not clear. (2) Indications of the nature and magnitude of structural perturbation's that attend introduction of fluorine into a protein are not available. (3) There is no reliable way to predict the direction or magnitude of protein-induced fluorine chemical shift effects. (4) Further development of experimental approaches that would reveal details of the structure and dynamical properties of the network of hydrogen atoms that surround a fluorine nucleus in a protein is needed; such information could be of value in defining the nature of a binding site that the design of new pharmacological agents could be assisted. The work proposed hopes to explore these considerations through experimental and computational studies of two representative serine proteases and isoforms of carbonic anhydrase. These systems were chosen so that the above aspects of the fluorine-labeling approach are illuminated while new information about proteins of continuing interest is produced.